Final Predictions in the Presidential Race

As expected, the polls are starting to tighten. The media is mostly still talking as though a Barack Obama landslide is still on track. Last week at this time, it was, but mountains of anecdotal evidence shows John McCain is again within striking distance, and the specifics of this race could yet turn it into a nailbiter, if not an outright McCain upset. Not only does the obvious issue--race--still loom large, but are the polls accurate in their weighting for larger numbers of Democratic voters than in 2004? Are enough of these "newly registered voters" gonna show up at the polls to change the game? Will the GOP robocalls calling Obama a terrorist change enough minds in these pink states? Will Osama bin Laden surface with a video endorsing Barack Obama so that he can once again get another hawkish Republican recruiting tool elected? Will the Michelle Obama "whitey" tape get unveiled? Any one of these things will benefit McCain.

Beyond that, I think Obama's decision to purchase a half-hour of primetime TV time in the final week of the campaign is a HUGE and potentially fatal mistake. Obama has been felled in the past by "overexposure" (New Hampshire primary, the speech in Berlin) while McCain benefits tremendously from the David vs. Goliath narrative, railing against Moneybags Obama and his attempt to simultaneously buy an election and disrupt the World Series! Last year, Alex Rodriguez interrupted the World Series with his drama....and it went over like a fart in church. I would have thought the Obama campaign would have learned their lesson about overreach at this point, but apparently not. The best spin I can put on this move is that they're hedging their bets and offering Obama a venue to discount last-minute Republican smears against him, but if Obama is doing that six days before the election, his campaign's in trouble anyway.

The point to all of this is that I think the race will close tight. I give Obama 60-40 odds of victory, but will be less surprised than most if McCain rises from the ashes again. Keep in mind that pre-election polls showed Bill Clinton leading Bob Dole by anywhere from 15 to 20 points....but he won by only eight. On that front, John McCain probably shouldn't mind all of the comparisons to Bob Dole.

Anyway, I always make final election predictions a few weeks before the election because I consider it cheating if I do it two days before the actual vote. I'll carry on that tradition this year and make my calls for all 50 states below...

Alabama--This will be McCain's strongest state in his strongest region--the South--and his coattails could hurt Democratic chances in two competitive House races. McCain by 28.

Alaska--Recent polls indicate Obama could still overperform Gore and Kerry in Alaska even with hometown girl Sarah Palin on the ticket. I still think it'll be a blowout though. McCain by 25.

Arizona--Why aren't there ANY polls on Arizona in recent months?!?! It's very frustrating because older polls shows Obama just outside of striking distance. Is the media embarrassed for McCain that he might get beaten in his home state or are they too lazy and clueless to recognize it MIGHT be competitive and actually poll the damn place? I suspect the latter, and the Obama campaign's inability to justify expanding its campaign there is likely to make a McCain blowout a self-fulfilling prophesy. McCain by 17.

Arkansas--Here's another state where the lack of Obama presence is troubling, not to mention a little insulting. At this point, Obama is poised to run ads in Kentucky, but not Arkansas? Huh? While it's likely true that there are few states where race looms as large as it would in Arkansas, it's a cheap market and one of the most Democratic states in America. Obama could win there if his campaign bothered trying, but as of now it looks like they're forfeiting it to McCain while tilting at windmills in Georgia. McCain by 12.

California--Most polls show an Obama blowout more pronounced than either Gore or Kerry pulled out, but I remain skeptical of any polls that measure large numbers of Latinos. Most Latinos can't legally vote, but can participate in polls, meaning it's likely that Obama support is overrepresented in California polls. Nonetheless, the west coast, even excluding Latinos, seems to be taking an even more substantial left turn this cycle, so I'll give Obama the benefit of the doubt and assume the polls are close to right. Obama by 13.

Colorado--Another state where Obama seemed like a shoo-in two weeks ago but where polls have largely stagnated with a modest Obama lead. Obama almost has to win in this state where he's led for months now if he wants to get to 270 electoral votes, and I suspect he probably will at this point, but I hope he's planning at least one more rally there between now and election day because it's not out of the question for McCain to pull it out. Obama by 2.

Connecticut--The northeast is a region where Sarah Palin has probably really drug down John McCain. Had McCain made a sane veep selection, he could have conceivably been competitive in the region, but it seems now as if he's likely to badly underperform Bush in 2004. Obama by 15.

Delaware--The media often overestimates Democratic strength in tiny Delaware. Kerry, after all, won it by only seven points in 2004. Nonetheless, with favorite son Joe Biden on the Democratic ticket this year, Delaware should be one of Obama's strongest states. Obama by 19.

District of Columbia--Obama should be able to pull off the seemingly impossible and actually grow the margins in the District of Columbia this year, despite its already being more Democratic than any state or even any COUNTY anywhere else in America. Obama by 83.

Florida--Polls always overestimate Democratic strength in Florida, and my sense is that the Obama bubble of the last two weeks is starting to deflate. Expect the trendline to move more and more to McCain in Florida in the next 18 days, and for McCain to win it fairly comfortably. McCain by 4.

Georgia--I'll say this much. Obama has made Georgia closer than I ever would have expected, but even at the height of the Obama bubble, he's never come closer than six points in any poll. I realize the campaign is now trying to boost black turnout and provide coattails for Senate candidate Jim Martin, but I still think it's a big waste of money in a very expensive advertising market, and that voters will break sharply Republican on election day. McCain by 13.

Hawaii--There has been a dearth of polling from Hawaii, much like there was four years ago until John Kerry had a ninth-inning scare there. This year, I'm not at all nervous about a repeat of that performance, if for no other reason that Obama is likely as close as Hawaii will get to having a native in the Oval Office. Obama by 26.

Idaho--Most of the Republican strongholds in the Rocky Mountain West have shown surprising strength for Obama this year compared to Gore and Kerry in elections past, but not Idaho. The little bit of polling I've seen from the state indicates that the militia men of Idaho have not taken to Obama. The needle will probably move a little in Obama's direction compared to 2000 and 2004, but not by enough to be measurable or indicate any kind of trendline. McCain by 36.

Illinois--Pretty safe bet that Obama wins commandingly in his home state, although not by the better-than-two-to-one margin he scored against Alan Keyes in the 2004 Senate race. Obama by 22.

Indiana--Even amidst the Obama bubble, McCain was starting to break away in Indiana. Now he's on the airwaves there (he wasn't before), and I think it'll become clear next week that it's slipped away from Obama. Even so, I think it was worth Obama's time to make such a serious effort there. McCain by 8.

Iowa--I remain skeptical of all the polls showing Obama with such a commanding lead in Iowa, but it will almost certainly be a blue state this year. In no other state is McCain's long-standing hostility to corn farmers in general and ethanol in particular gonna be more of an Achille's heel for him. Obama by 6.

Kansas--It's intriguing how the Northern Plains seem to have been far more accepting of switching teams to support Obama than the Southern Plains. The line generally seems to have been drawn between Nebraska and Kansas, with the former considerably less pro-McCain than it was pro-Bush, while the latter hasn't moved much despite Obama's background in the state. Then again, it shouldn't be that much of a surprise. Religion plays a much larger role in the conservatism of the Southern Plains than the Northern Plains. McCain by 20.

Kentucky--I realize Obama's recently announced plans to run ads in Kentucky have two goals in mind....because ads in Louisville filter into southern Indiana, and to help boost turnout for Democratic Senate candidate Bruce Lunsford. Even so, I'm not sure Kentucky is a state where an assist from Obama is what downballot Democrats would want. It strikes me as the state where racism is most likely to seriously affect the race, given the small numbers of non-whites compared to more traditional Old Confederacy bastions like Mississippi and Alabama. Perhaps I'm not giving Appalachians enough credit, but I still think Kentucky will be one of the few states where McCain will overperform Bush's 2004 numbers. McCain by 23.

Louisiana--Another state where McCain will overperform Bush. Louisiana would likely welcome a new Hurricane Katrina every day before they'd welcome a President Barack Obama. McCain by 19.

Maine--There was speculation two weeks ago that Maine could become competitive and that McCain could take the state's northern Congressional district's electoral vote. No sooner was that hypothetical thrown out there than the RNC announced it was pulling its ads from Maine. That certainly assuaged my fears about the state. Obama by 8.

Maryland--Here's a state the seems demographically perfect for Barack Obama with large numbers of blacks and white professionals. Even some of the rural regions of Maryland strike me as more competitive than usual this cycle. Obama by 16.

Massachusetts--I originally had concerns that initial Bay State skepticism towards Barack Obama would result in a seriously disappointing Democratic margin there, but like Connecticut, I think the obscenity against civilization that is Sarah Palin has helped Obama seem more acceptable by default even to many of the most skeptical Irish Catholic "waitress grandma" Hillary voters from Massachusetts. Obama by 23.

Michigan--I continue to be amazed that McCain picked the high-profile state of Michigan, expected to be THE swing state of the 2008 cycle, as the place to stage an early October retreat. What a horrible campaign decision. A retreat from Minnesota or Iowa would have been much less consequential in the minds of campaign handicappers than Michigan, and would have accomplished the same goal of freeing up resources for more winnable venues. It now seems inevitable that Obama will win Michigan, but I still anticipate it to be closer than what the McCain campaign predicted where they cut and run. Obama by 4.

Minnesota--In no other state has polling been more erratic this cycle than in my home state of Minnesota. For every poll showing a double-digit Obama landslide, another poll comes out showing it to be a one-point race. The truth almost certainly lies in the middle, but I'll be more intrigued than usual to see the returns roll in from Minnesota on election night. Obama by 6.

Mississippi--Will black turnout be large enough to significantly shrink GOP margins in 2008? Probably by some, but whites will respond by supporting McCain even more than they did Bush, so the net difference is unlikely to be substantial. McCain by 13.

Missouri--Polling has been somewhat erratic here too, but with a definite trendline to Obama in the last couple weeks. Nonetheless, I think it will close pretty strongly for McCain, as will the nation at large. McCain by 4.

Montana--First it was close in Montana, then McCain pulled away after the GOP convention, and I anticipate polls released in the days ahead show the race to be close again. With 30-second TV ads sold as cheap as $53 in Montana, it was a no-brainer for Obama to establish a campaign presence here, and to keep it here even after polls showed the state moving towards McCain. With that said, I suspect things to move back to McCain in the closing days of the race. McCain by 8.

Nebraska--Every indication is that Obama's attempt to steal the Omaha electoral vote from the McCain column is not working. Obama will most certainly shrink GOP margins in both Omaha and the state at large, but not by enough to alter the landscape in one of the nation's reddest states. McCain by 23.

Nevada--This is an easy state for Democrats to make close, but a tough state for Democrats to win. It's hard to accept the polls at face value since so many would-be Obama voters are working-class noncitizen Latinos who can't vote. Still, Obama has made wise decisions to campaign in conservative rural bastions like Elko in northeastern Nevada in attempt to hold down the landslide GOP margins that helped elect Bush twice in the state. Furthermore, a localized poll in Washoe County (Reno) showed Obama leading there. If Obama wins Washoe County, it's hard to see how he loses statewide. With that in mind, I'll carry on my quadrennial tradition of predicting a narrow Democratic victory in Nevada. Obama by 1.

New Hampshire--The trendline has moved decidedly towards Obama in the Granite State in the last couple of weeks, but the same was true in October 2004 when it looked like John Kerry was a sure thing in NH yet only won by one point on election night. I expect a similar photo finish this year. Obama by 2.

New Jersey--Obama will probably finish somewhere in between Gore's landslide margin in 2000 and Kerry's fairly weak seven-point victory in 2004, although New Jersey closes well for Democrats and Obama could conceivably score a Gore-sized landslide here. Obama by 12.

New Mexico--Like California and Nevada, noncitizen Latinos are probably artificially inflating Obama's margins in New Mexico polls, just as was apparently the case in 2004 when it looked like Kerry would win there. Obama should win, but it probably will be close. Obama by 3.

New York--Obama was poised to have a fairly weak victory in New York just two months ago, but since then two things have happened. The New York-centric financial crisis and McCain's pick of the noxious Sarah Palin as his veep. Beyond that, the New York Republican Party has completely unraveled. Obama should win by traditional margins in New York, if not a little better. Obama by 22.

North Carolina--Never could I have imagined that North Carolina would be in play this election cycle, and just the idea of it being colored blue on TV screens across America on election night is a development that can only be described as stunning. But let's keep it real here. Traditional sleazy GOP tactics of character assassination, combined with a more significant presence by the McCain campaign, are almost certain to take their toll on enough conservative Southerners to swing this election back to McCain in the closing days. McCain by 5.

North Dakota--Like Montana, the pendulum of competitiveness has swung back and forth dizzyingly in North Dakota, with new momentum for Obama right now. Now, Obama's going back on the airwaves in North Dakota. I'd like to be bullish here, but still anticipate enough voters revert to traditional voting patterns to give McCain a moderately comfortable victory. McCain by 6.

Ohio--It's striking that even amidst the Obama bubble of the last three weeks, the needle has only moved a few points in Ohio. The average poll margin is still within a couple of points. That tells me Obama cannot sustain the character assault he's about to receive by the robocalls and 527s. Even last week I would have predicted Ohio to go McCain. Now I'm very confident of it. McCain by 3.

Oklahoma--In America's purest enclave of hopeless "values voter" kamikazes, expect John McCain to get his strongest margin in the country. McCain by 42.

Oregon--All polls suggest Obama surging to what seems like an improbable lead in a state where Democrats usually win, but by soft margins (Kerry won by only four points). I'm not buying into the idea of a double-digit Obama win, but I do think Obama will easily overperform Kerry. Obama by 8.

Pennsylvania--The polls in Pennsylvania have been artificially inflated for Obama for the last two weeks. Obama will not score Bob Casey-esque margins in the state, and the racist robocalls and character assassinations will have a receptive audience with many older voters out in John Murtha's corner of the state. Obama will win, but it'll close strong for McCain, with Obama's margins being almost entirely the product of huge margins in suburban Philadelphia. Obama by 4.

Rhode Island--Like Massachusetts, it's likely that Obama's margins have hardened in the last month or so and it should be a Dem landslide. Obama by 25.

South Carolina--Whatever happens in North Carolina, don't expect much to change in South Carolina, where the demographics remain every bit as inhospitable as they were 20 years ago. Sure, there will be a boost of turnout among black voters that will help Obama a little, but it will be almost entirely offset by whites voting even stronger for McCain than they did for Bush. McCain by 15.

South Dakota--It seems hard to believe that Obama could be doing as well as he now appears to be in North Dakota without some softening of McCain numbers in its neighbor to the south, but there are no polls to vindicate that theory. Thus, I'll merely speculate that Obama has closed the gap, but that McCain will close well in the end. McCain by 13.

Tennessee--I must admit to being a little surprised that Tennessee appears poised to go as Republican as now seems likely given that African-American Senate candidate Harold Ford came within three points in Tennessee two years ago. Then again, Ford was a conservative and Obama is a liberal, so race might not be as big of a factor as we suspect. Tennessee is simply becoming a very conservative state despite its long-standing reputation as an island of moderation in the conservative South. McCain by 22.

Texas--Some polls show it modestly close, but again we have to assume many noncitizen Latinos who can't vote are being polled for Obama. Native son Bush isn't on the ballot, but even so it's likely that much of Texas will vote like Oklahoma this year, that is to say more Republican even than usual. This will be offset some by Democratic gains in urban areas, but I still don't see a 10-point race in Texas as being likely. McCain by 15.

Utah--I have little doubt that GOP margins will be less extreme than usual this year in Utah, but I do think they'll be surpassed by Oklahoma as the reddest state this election cycle. McCain by 40.

Vermont--I can never quite understand the culture of Vermont....how the most rural state in America can arguably be its most liberal. Either way, it could be very well be Obama's best state this year. Obama by 25.

Virginia--After months of skepticism, I've now seen enough polling data to agree that Obama has a better chance than not of winning Virginia. Margins will tighten and it could again become close, but the kicker for me was a localized poll for exurban Prince William County in northern Virginia. Tim Kaine narrowly lost it in the 2005 gubernatorial race while Jim Webb narrowly won it in the 2006 Senate race. As for Barack Obama, the poll said he's leading in Prince William County by eight points. Even if that margin shrinks by half in the closing days of the campaign, it seems almost impossible for the arithmatic to work in McCain's favor. If Obama wins Virginia, he's President. Game, set, and match. Obama by 2.

Washington--Obama had a bit of a scare in Washington in early September as polls narrowed, but he's back on track now and seems pretty likely to win by an even more comfortable margin than John Kerry's seven-point margin of 2004. Obama by 9.

West Virginia--My suspicions from last month that the state was much closer than anyone previously expected have to come to pass. It'll undoubtedly close for McCain again, but it's rather refreshing to see more Mountain State voters were willing to give Obama a hearing than was ever expected in the May primary blowout. McCain by 8.

Wisconsin--I'm a little surprised the RNC is poised to leave Wisconsin. The Quinnipiac Poll showing Obama leading by 17 there was a definite outlier, and I still see the state closing for McCain. He was probably never gonna win the Badger State this year, but had he stayed, he probably could have made it close. Obama by 5.

Wyoming--All polling evidence suggests McCain is dramatically underperforming Bush in this GOP stronghold. It'll still be a blowout, but no longer in the same league as Idaho or Utah, or now Oklahoma. McCain by 29.

My final tally is 291 electoral votes for Obama and 247 for McCain. Obama is right to warn against overconfidence as his poll position today has changed dramatically from where it was three days ago. Even during the time it took me to write this, new tracking polls were released showing Obama's lead continuing to crater. Apparently, any point where we are two days removed from disastrous economic headlines means the traditional trivialities of Republican campaigns will stick. As sick as it may sound, Obama desperately needs more bad economic data or this thing could yet slip away from him. Nonetheless, these are my final predictions and with the exception of Nevada, seem fairly hard in their likelihood of holding firm in the next 18 days.

I welcome either cut-and-pastes or links to the websites of other who have made their own predictions.